Event Rule on MACAddress - Custom Script fail with error

[stefanoparis-mgh]

Deployment Type

Self-hosted

NetBox Version

v4.4.0

Python Version

3.11

Steps to Reproduce

1. Define simple custom script that log a row (script attached)
2. Create Event Rule on MACAddress Create/Update that execute previously created script
3. Edit MACAddress

Script can be executed manually without error. In version 4.3.7 script worked correctly

Test.py

Expected Behavior

Script get executed with a simple log row

Observed Behavior

Error raised:
AttributeError("'NoneType' object has no attribute 'id'")

[stefanoparis-mgh]

Same with Event Rule on Journal Entry or Device created/updated

[miked314]

I'm getting same error. I took similar steps as @stefanoparis-mgh to confirm it wasn't my script.
here is output from netbox-rq

Sep 09 22:33:10  python3[1290826]: 17:33:10 default: handle(data={'id': 3280, 'url': '/api/dcim/devices/3280/', 'display_url': '/dcim/device..., job=<Job: test sync>) (2f1be4ee-3454-405c-be8d-44a5d7cb8add)
Sep 09 22:33:10  python3[1292870]: 17:33:10 Successfully completed handle(data={'id': 3280, 'url': '/api/dcim/devices/3280/', 'display_url': '/dcim/device..., job=<Job: test sync>) job in 0:00:00.031599s on worker aae833a4574b4c96809eeb964b164be7
Sep 09 22:33:10  python3[1292870]: 17:33:10 default: Job OK (2f1be4ee-3454-405c-be8d-44a5d7cb8add)
Sep 09 22:33:10  python3[1292870]: 17:33:10 Result is kept for 500 seconds

Error raised:

AttributeError("'NoneType' object has no attribute 'id'")

[rfdrake]

I was able to duplicate this:

In the Job Log I saw this:

In the logs for netbox-worker it shows that the job completed OK:.

netbox-worker-1  | 20:18:59 default: handle(data={'id': 155, 'url': '/api/dcim/devices/155/', 'display_url': '/dcim/devices/..., job=<Job: Test Script>) (b856a589-44af-4a70-970a-7f89e3a42931)
netbox-worker-1  | 20:19:00 Successfully completed handle(data={'id': 155, 'url': '/api/dcim/devices/155/', 'display_url': '/dcim/devices/..., job=<Job: Test Script>) job in 0:00:00.181012s on worker a6b02f245c97439baba44b8e1a26f623
netbox-worker-1  | 20:19:00 default: Job OK (b856a589-44af-4a70-970a-7f89e3a42931)

Simulating it in nbshell:

from extras.models import EventRule
from extras.jobs import ScriptJob
from users.models import User

er = EventRule.objects.get(name="test")
script = er.action_object.python_class()
test_job = ScriptJob.enqueue(instance=er.action_object, name=script.name, user=User.objects.get(username="admin"), data=None)

When I run this I can see the event processed with netbox-worker. I see the error in the Jobs console.

So I dug into ScriptJobs.run(). From what I can see, this is probably the culprit:

        self.logger.debug(f"Request ID: {request.id}")

I think in this particular case the request is None. I think this happens in two scenarios, one being events fired from event rules, and the other being scripts run on the CLI with runscript.

In runscript's case, it should still be working because it fakes the request, which gives it an id.

        job = ScriptJob.enqueue(
            instance=script_obj,
            user=user,
            immediate=True,
            data=form.cleaned_data,
            request=NetBoxFakeRequest({
                'META': {},
                'POST': data,
                'GET': {},
                'FILES': {},
                'user': user,
                'path': '',
                'id': uuid.uuid4()
            }),
            commit=commit,

So the simple fix would be edit jobs.py and change that debug line to make request.id optional (maybe move it two lines up under "if request:"

[miked314]

@rfdrake - Thank you!